I'd put every webservice in a physical server in a basement somewhere and be done with it.
One of the benefits of working in healthcare is that if it's not on-premesis, it's a non-starter for many organizations.
My company's paranoia of cloud services has paid off several times in terms of security in the last few years, even if the cost in dollars was higher. At this very moment, a vendor is going through a ransomware attack and its SaaS service unavailable. Our locally-hosted version is unaffected.
We only rent other people's computers (cough "cloud" cough) when there is absolutely no alternative, including doing without. On the other hand, it's locked me out of a number of programs and services I'd like to use.
Still, startups have to learn that if you're serious about being in enterprise, having a locally hosted option opens doors. Very lucrative doors.
I hope no startups learn that lesson, but instead hone their products with quick turnaround from users and then when they have achieved marked fit then offer to sell their product to enterprise companies for a massive fee. And by package I mean they deliver some servers with Kubernetes on it and run the same software they run in the cloud.
If a startup tries to sell to the enterprise they are likely to die before they make enough in sales and enterprise concerns (such as on prem >> price) aren't usually shared with normal businesses.
Not saying you can't make a fortune selling overpriced products to large companies (hello Oracle) but it isn't a game for a start up.
> Still, startups have to learn that if you're serious about being in enterprise, having a locally hosted option opens doors. Very lucrative doors.
Can you explain why that is?
I do agree with your point, having been exposed to running openstack on a custom infrastructure for my job has really taught me how much $ you can save if you do things yourself over say aws and that once I picked up the basics of compute, volumes, networking and security. My understanding increased.
I sometimes would find it difficult to keep up with all the aws lingo but doing everything myself has been a really good lesson in keeping things small and lean.
Usually has to do with security requirements, often dictated by the industry they're in. Health care and military contracting are both huge, and are both areas that large companies tend to end up serving even if they didn't set out to do that, with the result that lots of enterprises need to be able to satisfy a variety of security check-lists, most of which are easier to deal with if a service is self-hosted. This can include things like "must guarantee no traffic containing X goes over an insecure network" or "must guarantee no data ever transits or is stored in [LIST OF COUNTRIES]". These directives can be in conflict for different orgs, too, so it's not something you can just fix one time in your hosted version and appeal to everyone.
One of the benefits of working in healthcare is that if it's not on-premesis, it's a non-starter for many organizations.
My company's paranoia of cloud services has paid off several times in terms of security in the last few years, even if the cost in dollars was higher. At this very moment, a vendor is going through a ransomware attack and its SaaS service unavailable. Our locally-hosted version is unaffected.
We only rent other people's computers (cough "cloud" cough) when there is absolutely no alternative, including doing without. On the other hand, it's locked me out of a number of programs and services I'd like to use.
Still, startups have to learn that if you're serious about being in enterprise, having a locally hosted option opens doors. Very lucrative doors.